Send to printer »

Tutorials : Apr 1, 2010 (Vol. 30, No. 7)

Improving Productivity in Bioreactors

Process Aims to Boost Cell Culture Volumes
  • Rolf Zouwenga
  • ,
  • Alberto D'Avino, Ph.D.
  • ,
  • Gerben Zijlstra, Ph.D.

Biopharmaceutical companies developing recombinant therapeutics have faced several challenges over the last two decades. In particular, product demand has increased so dramatically that a capacity shortage has been anticipated. As a result, intensive capital efforts have been made to raise manufacturing capacities.

As a consequence of the high capital intensity, productivity improvements were required in order to manufacture affordable drugs. Just ten years ago, product titers of 0.5 g/L were regarded as significant. Today cell-culture titers in the range of 2–5 g/L have become the new baseline, and titers exceeding 5.0 g/L have been reported.

Together with advances in cell-line engineering, a major driver for these improvements has been optimization of process efficiencies, particularly upstream.

Bioreactors are commonly operated in two different modes: fed-batch and perfusion. In classical fed-batch culture, cells are supplemented with a concentrated nutrient solution, which results in improved growth and productivity. However, the accumulation of toxic metabolites and/or osmolality changes eventually leads to a declining culture. Cells are harvested along with concentrated product at the end of the process.

In perfusion culture, fresh medium is continuously added while exhausted medium-containing product is removed and cells are retained in the bioreactor by a retention device. Perfusion systems generally allow higher cell densities and volumetric productivities to be achieved. A major disadvantage of perfusion cultures, however, is that large volumes of diluted product are generated. Product recovery and purification from such bountiful harvests imposes significant challenges downstream.

DSM Biologics has developed an intensified process (high cell density and high titer) for recombinant protein and antibody production from mammalian cells called XD®. The system harnesses the advantages of fed-batch and perfusion systems into a hybrid process that produces a single concentrated batch at the end of roughly two weeks of cell culture. 

The XD process is designed to continuously supply fresh nutrients and constantly remove toxic metabolites while both cells and product are retained within the bioreactor (Figure 1). The process makes use of a separation system such as a hollow fiber filter operated in cross flow filtration mode. The molecular weight cut-off of the filter is optimized in order to retain the product within the bioreactor and, at the same time, allow removal of low molecular weight metabolites.

One option is to use an alternating tangential flow device, usually a hollow fiber module, along with XD to avoid filter clogging. The module is connected to the bioreactor by a single port. The cell suspension is pumped into the filtration module and back to the bioreactor by a low-shear pump. In this way, efficient tangential flow under low shear conditions as well as a frequent back-flush of the membrane is enabled. This reduces filter clogging and prolongs the life of the filter.

The XD process was designed to be virtually clone- and product-independent. It  has been tested for multiple mammalian cell lines (e.g., CHO, PER.C6® cell line, and myeloma). XD has produced viable cell densities as high as 150–200 x 106 cells/mL and IgG productivities of over 20 g/L. A titer of 27 g/L was reported using an IgG-producing PER.C6 clone.

The XD process ensures consistent process conditions, which improves product quality. Comparison of quality data for an IgG using a PER.C6 clone showed no differences between fed-batch or the XD process for glycosylation pattern. All other relevant assays showed similar results.

Case Studies

Early trials applying the XD process for an IgG-producing CHO clone resulted in a viable cell density of approximately 60 x 106 cells/mL and 12.6 g/L of antibody. In addition, it was found that by shortening the cultivation period to 15 days, operational costs were reduced and risks associated with long processes (e.g., contamination and technical failures) diminished.

Another advantage of XD is that as a result of production improvements, the overall capacity of a manufacturing plant may be reduced if the same output of product is desired, decreasing plant footprint, equipment, and facility-related investments. Also, because product is collected inside the bioreactor, only a small and concentrated batch is harvested, thus eliminating the need to collect and purify large, multiple volume harvest batches.

In another experiment, a CHO clone engineered to produce and secrete a fusion protein was tested. The original fed-batch process from the customer utilized commercially available, chemically defined medium and feeds, and reported a maximum product concentration of only 1.2 g/L in fed-batch mode.

By applying XD process technology, with the same basal medium, a biomass concentration of 130 x 106 cells/mL was attained (Figure 2). The high cell growth in the XD bioreactor translated into a 7.7-fold improvement in productivity (final product concentration of 9.2 g/L).

The benefits of boosting the productivity of the bioreactor are obvious. The original fed-batch process utilized a 400 L bioreactor that delivered 480 grams of recombinant product. Using XD process technology the same amount of product was obtained by using a 50 L working volume bioreactor. Even higher productivity was obtained in a second run using the same CHO clone.

Cells were at first adapted into a different chemically defined formulation. An XD process was then performed using the same medium. A cell density of 175 x 106 cells/mL was achieved on day 12 (Figure 3). Product concentration of 11.3 g/L was measured on the same day, corresponding to a 9.5-fold factor improvement of the fed-batch productivity, while keeping the same length of the original process (12 days).

XD process technology integrates cell culture and product retention. It is almost independent from the cell line.

XD process technology may also boost cell densities and the productivity of CHO clones making different products (monoclonal antibody and fusion protein). By boosting volumetric productivity of bioreactors to high levels, XD process technology is ideal for situations in which large amounts of product needed to be made quickly and with minimum development and investment. XD is also ideal for manufacturing products that are difficult to produce at high titers in traditional fed-batch systems.